PAC Verification of Statistical Algorithms
Abstract
Goldwasser et al. (2021) recently proposed the setting of PAC verification, where a hypothesis (machine learning model) that purportedly satisfies the agnostic PAC learning objective is verified using an interactive proof. In this paper we develop this notion further in a number of ways. First, we prove a lower bound of i.i.d.\ samples for PAC verification of hypothesis classes of VC dimension . Second, we present a protocol for PAC verification of unions of intervals over that improves upon their proposed protocol for that task, and matches our lower bound's dependence on . Third, we introduce a natural generalization of their definition to verification of general statistical algorithms, which is applicable to a wider variety of settings beyond agnostic PAC learning. Showcasing our proposed definition, our final result is a protocol for the verification of statistical query algorithms that satisfy a combinatorial constraint on their queries.
Keywords
Cite
@article{arxiv.2211.17096,
title = {PAC Verification of Statistical Algorithms},
author = {Saachi Mutreja and Jonathan Shafer},
journal= {arXiv preprint arXiv:2211.17096},
year = {2023}
}